llvm-6502/lib/Analysis
Tobias Grosser a10d598602 Instantiate DefaultDOTGraphTraits
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90133 91177308-0d34-0410-b5e6-96231b3b80d8
2009-11-30 12:38:13 +00:00
..
IPA
AliasAnalysis.cpp use the new isNoAlias method to simplify some code, only do an escaping check if 2009-11-23 16:46:41 +00:00
AliasAnalysisCounter.cpp
AliasAnalysisEvaluator.cpp
AliasDebugger.cpp
AliasSetTracker.cpp
Analysis.cpp
BasicAliasAnalysis.cpp move DecomposeGEPExpression out into ValueTracking.cpp 2009-11-26 17:12:50 +00:00
CaptureTracking.cpp
CFGPrinter.cpp
CMakeLists.txt
ConstantFolding.cpp Teach ConstantFolding to do a better job when folding gep(bitcast). 2009-11-29 21:40:55 +00:00
DbgInfoPrinter.cpp
DebugInfo.cpp Use StringRef (again) in DebugInfo interface. 2009-11-25 17:36:49 +00:00
DomPrinter.cpp Instantiate DefaultDOTGraphTraits 2009-11-30 12:38:13 +00:00
InlineCost.cpp
InstCount.cpp
InstructionSimplify.cpp factor some logic out of instcombine into a new SimplifyAddInst method. 2009-11-27 17:42:22 +00:00
Interval.cpp
IntervalPartition.cpp
IVUsers.cpp enable iv-users simplification by default 2009-11-23 23:25:54 +00:00
LazyValueInfo.cpp
LibCallAliasAnalysis.cpp
LibCallSemantics.cpp
LiveValues.cpp
LoopDependenceAnalysis.cpp
LoopInfo.cpp
LoopPass.cpp
Makefile
MemoryBuiltins.cpp
MemoryDependenceAnalysis.cpp Revert r90107, fixing test/Transforms/GVN/2009-11-29-ReverseMap.ll and the 2009-11-30 07:05:51 +00:00
PointerTracking.cpp
PostDominators.cpp
ProfileEstimatorPass.cpp
ProfileInfo.cpp
ProfileInfoLoader.cpp
ProfileInfoLoaderPass.cpp
ProfileVerifierPass.cpp
README.txt
ScalarEvolution.cpp Remove dead LLVMContext argument. 2009-11-23 03:26:09 +00:00
ScalarEvolutionAliasAnalysis.cpp
ScalarEvolutionExpander.cpp
SparsePropagation.cpp
Trace.cpp
ValueTracking.cpp implement a FIXME: limit the depth that DecomposeGEPExpression goes the same 2009-11-28 15:12:41 +00:00

Analysis Opportunities:

//===---------------------------------------------------------------------===//

In test/Transforms/LoopStrengthReduce/quadradic-exit-value.ll, the
ScalarEvolution expression for %r is this:

  {1,+,3,+,2}<loop>

Outside the loop, this could be evaluated simply as (%n * %n), however
ScalarEvolution currently evaluates it as

  (-2 + (2 * (trunc i65 (((zext i64 (-2 + %n) to i65) * (zext i64 (-1 + %n) to i65)) /u 2) to i64)) + (3 * %n))

In addition to being much more complicated, it involves i65 arithmetic,
which is very inefficient when expanded into code.

//===---------------------------------------------------------------------===//